Who Should Decide Procedural Fairness Questions?
The latest contributor to the growing literature on deference and procedural fairness is Adrian Vermeule in “Deference and Due Process“:
In the textbooks, procedural due process is a strictly judicial enterprise; although substantive entitlements are created by legislative and executive action, it is for courts to decide independently what process the Constitution requires. The notion that procedural due process might be committed primarily to the discretion of the agencies themselves is almost entirely absent from the academic literature.
The facts on the ground are very different. Thanks to converging strands of caselaw — partly involving due process, partly involving judicial deference to agency interpretation of procedural provisions in statutes, and partly involving the long shadow of Vermont Yankee v. NRDC — agencies themselves are now the primary front-line expositors and appliers of the cost-benefit balancing test of Mathews v. Eldridge. The courts for their part often defer, explicitly or implicitly, to agencies’ due process decisions.
I will defend this approach, and urge that it be made fully explicit. Rather than decide for themselves “what process is due,” courts should ask only whether the agency offered a rational justification for providing whatever process it did provide. Although the Mathews cost-benefit calculus would still supply the rule of decision, courts should merely review the application of that rule by agencies, and defer to reasonable agency decisions about the costs and benefits of procedural arrangements.
From the paper:
As has been true at least since Marbury v. Madison, however, a persistent occupational hazard of (some) lawyers is to skip lightly from the paramount status of the Constitution to the fallacious inference that courts must decide everything for themselves…A regime in which courts allow agencies to design procedures, subject to arbitrariness review, does not define protected interests out of existence. It recognizes them and then goes on to ask what institutional regime, what allocation of authority to design procedures, will best accommodate that interest as part of a larger system of interests that must be balanced, reconciled, and — at some resource frontier, assuming agencies do not have infinite time and infinite budgets — traded off against one another (p. 18).The agency’s problem is always ex ante and aggregated; its problem is to design and operate a procedural system over an array of cases, deciding at several margins simultaneously where and how to invest its limited resources in deciding what policies and legal rules to adopt. That sort of optimizing cannot be substance-independent; the agency will have to decide which sorts of errors are more costly, how many resources in can invest in which sorts of decisionmaking, and which sorts of proceedings or cases will yield the greatest return-on- investment for its policy goals (p. 28).[R]ather than plunging into procedural design themselves, courts should stop, look, listen, and then ask a logically antecedent question: is there any reason to think that the relevant agency will have systematically skewed incentives or make systematically distorted judgments when applying the Mathews calculus, such that the agency will systematically over-weight or under-weight particular factors? If there is no reason to think that the agency has systematically distorted incentives or is prone to systematic distortions of judgment, then there is no reason for judges to displace agency choices about the marginal costs and benefits of additional (or different) procedure in the execution of the agency’s program. Agencies may of course make random errors, but so too may judges — especially generalist judges deciding on the value of procedures under programs they only dimly understand (p. 30)
Note though that Vermeule (pp. 16-17) discounts the possibility that courts can defer by giving weight to decision-makers’ procedural choices even though retaining the final word on whether a particular procedure is fair. There is no mention of Skidmore deference in the paper. Canadian readers might usefully compare John Evans, “Fair’s Fair: Judging Administrative Procedures” (2015), 28 CJALP 111 (gated). See further here and here.
UPDATE: That this question is still in the balance in Canada is indicated by Bergeron v. Canada (Attorney General), 2015 FCA 160:
 The law concerning the standard of review for procedural fairness is currently unsettled. The unsettled nature of that law is shown by the Supreme Court’s recent decision in Mission Institution v. Khela, 2014 SCC 24 (CanLII),  1 S.C.R. 502, a procedural fairness case. In that decision, the Supreme Court declared, without elaboration, that the standard of review is correctness but just ten paragraphs later it found that some deference should be owed to the administrative decision-maker on some elements of the procedural decision: at paragraphs 79 and 89.
 Some cases of this Court have fastened onto the Supreme Court’s statement of correctness in Khela without noting the later words of deference: see, e.g., Air Canada v. Greenglass, 2014 FCA 288 (CanLII), 468 N.R. 184 at paragraph 26. Those cases have not referred to other cases of this Court that suggest that the standard is not purely correctness and that some deference can come to bear.
 For example, this Court has spoken of proceeding under correctness review but in a manner “respectful of the [decision-maker’s] choices” with “a degree of deference”: Re:Sound v. Fitness Industry Council of Canada, 2014 FCA 48 (CanLII), 455 N.R. 87 at paragraph 42. And this Court has also upheld reasonableness review, but on the basis of a variable margin of appreciation being afforded to the decision-maker (as explained above), sometimes a wide one and sometimes no margin at all: Forest Ethics Advocacy Association v. Canada (National Energy Board), 2014 FCA 245 (CanLII), 465 N.R. 152; and for a defence of this position see my dissenting reasons in Maritime Broadcasting System Limited v. Canadian Media Guild, 2014 FCA 59 (CanLII), 373 D.L.R. (4th) 167. And in this very context—whether procedural fairness was infringed by an insufficiently thorough investigation under the Canadian Human Rights Act—there is authority for the proposition that deference to the fact-based judgment of the Commissioner is warranted: Slattery v. Canada (Human Rights Commission), 1994 CanLII 3463 (FC),  2 F.C. 574, 73 F.T.R. 161 (T.D.) at paragraphs 55-56, aff’d (1996), 205 N.R. 383 (C.A.).
 One might also query whether a failure to investigate thoroughly under the Act is a procedural defect, triggering whatever standard of review applies to procedural matters. A decision based on a deficient investigation can be characterized as one that is not substantively acceptable or defensible because it is based on incomplete information, thereby triggering the standard of review for substantive defects governed by Dunsmuir, above. As was the case in Forest Ethics, above, the line between a procedural concern and a substantive concern can be a blurry one. As this Court explained in Forest Ethics, there is much to be said for the view that the same standard of review—reasonableness with variable margins of appreciation depending on the circumstances (as described earlier in these reasons)—should govern all administrative decisions.
 So what we have right now is a jurisprudential muddle. And now is not the time to try to resolve it. For one thing, we have not received submissions on the issue in this case. For another, with so many conflicting decisions, perhaps only a reasoned decision of the Supreme Court can provide clarity.
This content has been updated on July 22, 2015 at 11:10.